A New and Improved Experimental Design for the Discrete Choice Experiment Module of the EuroQol Valuation Technology Protocol.
Value Health
; 27(10): 1311-1317, 2024 Oct.
Article
em En
| MEDLINE
| ID: mdl-38977189
ABSTRACT
OBJECTIVES:
The EuroQol Valuation Technology (EQ-VT) protocol, which is used to value the EQ-5D-5L instrument, comprises a composite time trade-off and a discrete choice experiment (DCE) module. Despite significant limitations, the DCE module has not been updated since its inception in 2012. This study aimed to update the EQ-VT DCE design using state-of-the-art methods.METHODS:
DCE data from 19 EQ-5D-5L valuation studies were summarized using a Bayesian hierarchical meta-analysis model, which created the priors for our Bayesian efficient DCE design. This design comprised 20 subdesigns, each with 12 choice tasks, and included 2 levels that overlapped to reduce the complexity of the choice tasks. The relative efficiency and robustness of the new design were established by comparing the D-errors and minimal sample size requirements for the 19 within-sample and 7 out-of-sample countries with the previous DCE design.RESULTS:
The updated DCE design shows large reductions in the D-error by 20% and 22% for the 19 within-sample and 7 out-of-sample countries, respectively. Sample size requirements were also reduced, resulting in an average reduction of 45% for both the within and out-of-sample countries.CONCLUSIONS:
The updated DCE design outperforms the current EQ-VT design. Given its enhanced performance and reduced complexity, it is set to replace the existing DCE design in future EQ-5D-5L valuation studies using the EQ-VT protocol.Palavras-chave
Texto completo:
1
Bases de dados:
MEDLINE
Assunto principal:
Qualidade de Vida
/
Projetos de Pesquisa
/
Comportamento de Escolha
/
Teorema de Bayes
Limite:
Humans
Idioma:
En
Revista:
Value Health
/
Value health
/
Value in health
Assunto da revista:
FARMACOLOGIA
Ano de publicação:
2024
Tipo de documento:
Article